The present invention generally relates to apparatus and methods for recording data on film.
Many applications require the storage of very large quantities of data for long periods of time. One example is found in the production of motion pictures or movies. Traditionally, a movie was made by shooting an original camera negative (OCN), which was then edited by cutting and splicing operations. More recently, the use of digital special effects has introduced a requirement for some parts of the OCN to be “scanned” to convert each frame of film into a set of digital data, which represents the frame film image information. Similarly, when an old movie is to be restored using digital techniques it may be necessary to scan all of the film so as to obtain a digitized version of the entire movie (a digital film record).
The process of scanning a film to create a digital film record is expensive and time consuming, and each use of the OCN increases the possibility of damage. Hence, it would be advantageous to treat the digital film record itself as the archive of the content, rather than the traditional approach of creating three monochrome separations on film and archiving those.
In this regard, the scanning process generates very large quantities of data. Today, most film scanning is performed at a resolution of “2K”, meaning 2048 pixels horizontally by 1536 pixels vertically (or similar resolutions generating similar quantities of data). Generally each pixel is represented by a digital value for each of red, green, and blue representative signals, where each digital value has a precision of at least ten bits. This means that more than 11 MB (millions of bytes) of data is generated for each frame of a film. There are normally 24 frames in a second of film, yielding an effective data rate of approximately 300 MBps (megabytes per second). Thus, a 2-hour movie would be represented by more than 2 terabytes (TB) of data, where, for purposes of this description, each terabyte is defined herein as being equal to 1,000,000 MB.
However, data storage requirements will continue to increase. For example, scanning a film with a “4k” resolution, and 14-bit precision, generates about 66 MB for each frame of film. In addition, if digital techniques are used for all of the production it may be necessary to scan an OCN that exceeds by many times the duration of the final movie. These additional factors mean that the data storage requirements for a single movie may reach many tens of terabytes.
Such quantities of data are very difficult to handle and store. Current data tape mechanisms can transfer data to tape at a rate of approximately 50 MBps, and provide storage of approximately one half of a terabyte on a single tape. Using such a device, transferring a 2-hour movie at a 2K resolution to tape would require four tapes and take nearly twelve hours to complete.
Another problem with storage of data is longevity. Most magnetic media such as tape, and optical storage media such as CDs and DVDs, have expected lives of a few tens of years. These life spans are quite unacceptable for archival purposes. In comparison, the science of archiving film is well developed, and color film can be maintained in good condition for many tens of years, while monochrome separations on modern stock are expected to have a useful life of hundreds of years.
Film records are, in fact, quite different from most forms of records. For example, in magnetic recording or optical recording it may be possible to overwrite or destroy the record. In particular, in magnetic recording, elements are magnetized in a certain direction during the recording process. Clearly the same elements can be de-magnetized or re-magnetized by a re-application of the recording process. Similarly, while optical records generally take the form of physical indentations in a surface, or alterations in a dye layer which result in a localized change in the optical properties of the layer, that are not easily altered, it may be possible to overwrite, or at least to destroy, the optical record by a reapplication of the recording process.
In contrast, in a film record the application of light to photographic film (exposure) causes a latent image that is then subjected to a chemical “development” process so that the image is substantially reinforced. Unexposed emulsion is then removed by a process known as “fixing”. The combination of these processes yields a very robust record that can no longer be overwritten by re-exposure, or damaged by anything but extreme physical processes.
Although the movie industry has been used as an example, many other businesses create large amounts of data and have needs for archival storage of this data. It is interesting to note that the longevity of film records compared to other available storage mechanisms has been recognized in the data industry. Some companies such as Anacomp, Inc. offer services to businesses requiring long-term storage of computer data records. In one mechanism, data records are imaged as character displays on (for example) a cathode ray tube and recorded on film, usually 16 mm (milli-meter) or 105 mm microfiche. An example of this technology is described in U.S. Pat. No. 4,553,833, issued Nov. 19, 1985. In this patent, light emitted from a relatively large-sized array (such as a light emitting diode array) is focused through converging lenses to cause a relatively small-sized dot pattern to be projected on a film. While this approach yields records that can provide the desired degree of permanence, the data density is relatively low (perhaps 100 to 1000 bytes/mm2) and hence the technique is not suited to storage of very large data records.
It should be noted that binary data may also be recorded directly onto film as a pattern of black and white dots representing values of ones and zeroes. Generally some sophisticated coding scheme is used to improve the effective data density and to provide error correction capability to ensure robustness. These techniques have been used extensively for recording audio data onto the edge of a motion picture film. For example, U.S. Pat. No. 4,600,280, issued Jul. 15, 1986, describes a technique for recording a digital soundtrack on a film strip by exposing the film to modulated light from a light source. In one method disclosed therein an intermittent light beam (encoded with digital audio information) is scanned horizontally across the film, and the film is then advanced vertically and the scanning process repeated. This patent also describes that the light can be projected on the film through a linear array of solid state shutters or Bragg cell modulators.
Other examples of storing data on film are described in the following U.S. Patents. U.S. Pat. No. 4,461,552, issued Jul. 24, 1984, describes a method for photographically recording digital audio on motion picture film. U.S. Pat. No. 4,306,781, issued Dec. 22, 1981, describes recording a command data track on motion picture film, along with an unmodulated locator and several analog soundtracks. Similarly, both U.S. Pat. No. 4,659,198, issued Apr. 21, 1987, and U.S. Pat. No. 4,893,921, issued Jan. 16, 1990, describe a process for recording digital data along an edge portion of a strip of cinematographic film. And, U.S. Pat. No. 5,453,802, issued Sep. 26, 1995, discloses a method and apparatus for photographically recording digital audio signals, and a medium having digital audio signals photographically recorded thereon.
In general, the technology represented by the above-described patents provide a very robust signal that can survive the two printing processes generally necessary to generate a movie release print, and that is reasonably tolerant of minor damage that can occur with use, particularly to edges of a release print. Unfortunately, the recording density is below one kilobyte/mm2, which is too low for use in providing long-term archival storage of large amounts of data such as represented by, e.g., a digital film record.